Fighting fire risk with data
In Philadelphia, Fire Commissioner Lloyd Ayers collaborated with data analytics firm Buxton on the first U.S. pilot program to pinpoint high-fire-risk households throughout the city. The company worked with the Philadelphia Fire Department to develop a fire-vulnerability index that scored every household in city limits on its propensity for fire. Once identified, firefighters went door to door, block by block, to educate citizens with fire-safety checklists and to install free smoke alarms.
I recently discussed the investment with Ayers and how it benefits Philadelphia and its citizens.
Where do fire deaths occur in Philadelphia?
Most of the fire deaths happen in the disadvantaged neighborhoods, the poorer communities. So we try to focus on making sure they have life-safety devices, like smoke alarms, that absolutely save lives. The other thing is to ensure they have all of the education they need, as far as home escape planning, testing smoke alarms, and how they should react when they hear the alarm [sound]. We also want to make sure they understand how to stay safe in their environment by removing common causes of fires. So we have a home-fire safety checklist review with our citizens. We want to bring those deaths down.
How did you use data to target high-risk populations?
This is where we are now, with a fire vulnerability index. We started out by tracking all fires, looking at where they were occurring. We started out at the zip code level. Then we would overlay the zip code with the fire stations and find out which fire station or fire-safety district were having the fires—ensuring those officers were able to service folks with fire alarms. After that we got it down to the block level — with about 40 to 60 homes on each block — and made sure they were serviced. We would go out and knock on doors and give them literature, offer to install a free smoke alarm and review the fire safety checklist — such as overloading circuits, smoking in the house…. We have the opportunity to spend more time on education, so we can chase the fire out of the city.
How did you develop maps of high-risk target areas based on household data?
We worked with Buxton, a marketing-data firm, where they go into a more granular view of the population of our city or the household level. They look at what their purchase characteristics are, where they attended the schools. They look at how many people are in the home. They look at a various databases and bring those items together to build profiles… [or] a fire vulnerability index for each household. Now, we can focus efforts on certain households throughout the city and go to them before the fire starts.
What has been the ROI?
We are still in the process of measuring it. The large indicators are the fire deaths. We started out several years ago, and we had 52 fire deaths in 2005 to 2006. We now have 35 deaths, which is the lowest we have ever had in the city. It is a 43% drop in fire deaths, and we want to maintain that. Our goal is zero fire deaths, so we want to sustain what we have already gained in the decrease of fire loss and deaths as well as a decrease in ambulance and hospital care.
The fires are going down, and now we have fire on the run. Freedom from fire is what we want.